use dataset_9296, clear

sem ///
(moder_94 <- moder_92 ft_own_92 pstr_92 ft_opp_92) ///
(ft_own_94 <- ft_own_92 moder_92 pstr_92 ft_opp_92), ///
cov(e.moder_94*e.ft_own_94)
